Strategies to Edit Production Data
summary
Five concrete strategies progress from maintaining a google spreadsheet of raw SQL edits to running Python scripts locally, copying them to an EC2/Jenkins host, using a task runner like Jenkins, and finally shipping a custom Script Runner service. Examples show SQL mistakes, a real migration over ~50 million products, CPU and credential problems, how audit logs and code review get enforced, and how centralizing credentials, environment separation, parallel execution and preview modes make edits safer and reproducible.
AI-written summary · about
chapters (14)
- 0:00Intro
- 0:21Strategies to Edit Production Data
- 1:36Why We Edit Production Data
- 5:18How the strategy works What's great What's not great
- 5:40Develop a Review Process for Manual Edits
- 5:58UPDATE products SET name='julies-product' WHERE id=1
- 10:15Run Scripts Locally
- 10:543. Run the script
- 12:29Network disconnections
- 13:53Run Scripts on an Existing Server
- 14:493. SSH and run inside a session
- 18:18Use a task runner
- 19:45Code review and automated tests
- 22:13Build a Script Runner Service
related talks
30mStrategies to Edit Production Data
29mStrategies to Edit Production Data
26mStrategies to Edit Production Data
29mStrategies To Edit Production Data
25mOptimize your Data Pipeline without Rewriting it
29mHow to build a data pipeline without data
30mReliable Data Processing with Minimal Toil
20mExcellent Database Adventure: From Code Errors to Reliable Releases